INDEX
Negative Logits
by
1.45
in
1.37
The
1.22
t
1.20
of
1.17
p
1.17
م
1.15
1
1.11
在
1.11
was
1.10
POSITIVE LOGITS
ir
1.24
↵
1.12
-
0.92
be
0.90
ot
0.86
dàng
0.83
이나
0.82
it
0.82
에
0.81
میتوانید
0.80
Activations Density 0.002%
by
in
The
t
of
p
م
1
在
was
ir
↵
-
be
ot
dàng
이나
it
에
میتوانید